A brief history of Wi-Fi (802.11a to 802.11ax)

By Michael Spalter
September 2021
WiFi History

About the author

Michael Spalter

Michael Spalter

Michael Spalter has been a networking technician for over 30 years and has been the CEO of DrayTek in the UK since the company’s formation in 1997. He has written and lectured extensively on networking topics. If you have an idea for a blog or a topic you would like explored, please get in touch with us.

With the new 802.11ax wireless standard becoming mainstream, today I am looking back at the evolution of wireless LAN which is now 20 years old.

The ITU-T & The Wi-Fi Alliance

802.11 is the original wireless specification ratified by the ITU-T and international body, over 50 years old which has been responsible for many standards including DSL (ADSL, VDSL etc.), Optical networks, X.509 (Cybersecurity), H.264 (Video codecs). Over the decades, the ITU-T ratified standards for telex, fax, analogue modems, email, ISDN and even JPEG. The ITU-T itself does not invent technology or systems - it has committees that draw on the tech community and commercial entities to pull together the best ideas and technologies to make interoperable standards. hilst there may be a preference for open/free technologies, many ITU-T standards make use of patented technologies and in order to use an ITU-T you may still have to separately licence the technology or patents that a standard relies upon.

As these common standards often use patented technology, we regularly get patent trolls sending us threatening letters and demands for money because they believe that we are using a standard that contains 'their patent'. In every case, this has either been bogus or we had the correct/valid usage rights or licence, and they don't even check. That is not to say that patent breaches do not exist, but trawling expeditions and blanket accusations are the tools of trolls, not innovative creators who have been wronged. Of course, the whole topic of patents is a huge debate in itself - the balance of protecting a true invention, investment, and creativity against obvious, simple, or essential 'inventions' which include a low threshold of innovation.

The name "Wi-Fi" is a marketing term used by members of the Wi-Fi alliance - a group of equipment and chipset vendors formed to help with standards compliance and interoperability. The Wi-Fi alliance is also responsible for the WPA security standards (WPA, WPA2 and now WPA3) which add encryption to Wi-Fi. In 2018, the Wi-Fi alliance decided to make the naming of Wi-Fi standards more consumer-friendly, so they introduced aliases, "Wi-Fi 4", "Wi-Fi 5" and "Wi-Fi 6".

The pre-cursor to the 802.11 standards, the first system that we would recognise as a wireless LAN was called WaveLAN and it was developed for banking/cashier systems in 1990, though other systems had been in development and had limited use for years before that.


The original 802.11 standard was released in 1997 (hence also known as 802.11-1997) but it was poorly defined and had inconsistent interoperability between vendors as each tried to interpret ambiguous definitions. The specification was clarified in 1999.

Although we usually think of 802.11 as a wireless radio standard, it was designed to operate wirelessly over either radio (in the 2.4Mhz band, so actually microwave) or Infrared - between two devices with IR transceivers. If you were in IT in the early 21st century, you may recall PCs and other devices with Infra-Red LEDs and receivers. This was before Bluetooth and sending between PCs or a digital camera and your PC seemed miraculous however, 802.11-1997 was not widely used in either its Infra-Red or radio versions - the world was quite happy using cables and there were relatively few handheld or portable devices. Laptops were larger and users often didn't need continuous connectivity.


802.11b (1999)

The 802.11b standard (released in 2000) provided a raw data speed of 11Mb/s. To get that 11Mb/s, 802.11b used 20Mhz wide channels instead of the original 5Mhz of 802.11-1997, but with the band still only 82Mhz wide, and keeping the same channel numbering scheme, it meant that, channel frequencies overlapped.

In 1999, wireless LAN was rarely a standard feature on laptop or desktop PCs. It was a costly addition, typically requiring a PCMCIA card interface or ISA card (for desktop PCs). Most companies and individuals still did not have Internet access, laptops were not small or light enough to be used whilst moving and peripherals like modems or printers were not wireless.

Band Congestion

Whilst 802.11b used 20Mhz wide channels, later wireless standards could even wider channels (40 or 80Mhz) - each channel taking up multiple adjacent channels, like a very wide truck straddling multiple lanes on the highway. This diagram shows how channels overlap with multiple adjacent channels:

WLAN Channels

Collisions and interference will occur if your wireless device is within range of someone else's and their channel overlaps with yours.   The proliferation of Wi-Fi means that, except in very rural environments, you are always in range of many other wireless access points and you cannot avoid overlapping with many neighbours at once. 
Here's a screenshot from a Wi-Fi base that lists all other nearby access points - there are lots, but by checking which channels are in use, you can try to select the least congested channels with the weakest other signals (the RSSI is the Received Signal Strength Indicator - the estimated power level). Many wireless routers or access points can also be set to do this automatically.

AP Discovery

The upshot is that whatever speeds are theoretically possible on the wireless standards you are using - once you account for the physical environment, distances, other devices using the network and other nearby Wi-Fi networks -  you have a lot of congestion in a relatively small band. 

802.11a (1999)

802.11a was the first Wi-Fi protocol to use OFDM (Orthogonal Frequency Division Multiplexing) which allowed much more efficient use of the bandwidth and therefore a significantly higher raw data rate of 54Mb/s (about 20Mb/s throughput in practice). OFDM uses a longer signal time than the CCK/DSSS methods of 802.11b - i.e., each symbol is held for longer, giving the receiver more opportunity to receive a poor signal or where there is interference, making it more robust.   OFDM is, in my opinion, a very elegant method to reduce adjacent carrier and other signal interference. The 'orthogonal' part means that it uses trigonometric transformations of waveforms. If that sounds complicated or dull, see my article on 802.11ax where I explain OFDM in detail.

Although 802.11a was introduced (as a standard) in 1999, I have listed it after 802.11b because it was not available in products until much later. Chipset vendors took some time to develop products supporting the more complex OFDM and support for the 5Ghz band.

One significant difference between 802.11a and the previous and next three standards was that it used the 5Ghz band, initially a 200Mhz wide allocation which was more than double the bandwidth of the 2.4Mhz band. That allowed more adjacent wireless bases to avoid overlapping channels however the 5Ghz band, being a higher frequency, had a shorter range. As wireless LAN was rare in 1999, there was not actually any shortage of free channels or the same congestion that we have today.

802.11g (2003)

802.11g was very similar to 802.11a, using OFDM and providing 54Mb/s raw capacity but operated back in the more common 2.4Ghz band.  As 802.11b was the most popular standard by now, this meant that and 802.11g products could retain backward compatibility with those existing networks, whilst supporting the new standard.   If a device needs to support two bands, it requires two radio transceivers - adding significant cost.  Therefore, the adoption of 802.11g was much greater than 802.11a and as 802.11b had already become popular, the demand for greater speed was high.  In the rush to be 'first to market' many vendors released draft 802.11g products which were incompatible with the final specification, and therefore other vendors' products (this was repeated with 802.11n).  There were also proprietary variations of 802.11g such as Atheros's "Super G" which bonded two channels to provide a raw channel speed of 108Mb/s.

802.11n (2007) aka Wi-Fi 4

802.11n was released in 2007 and was retrospectively given the alias Wi-Fi 4 by the Wi-Fi alliance in 2018.  802.11n increased the raw data rate up to a maximum of 150Mb/s (from 802.11g's 54Mb/s) but also used multiple spatial streams (see 'MIMO' below) and channel widths up to 40Mhz to give a maximum raw data rate of 600Mbs/ (across all streams).  As previously, the actual performance would depend on the environment. A single stream client would still only have access to one 150Mb/s channel, assuming that there wasn't congestion necessitating the need to drop to a 20Mhz channel bandwidth and that you were able to use the shortest guard interval (400ns).   The 802.11n specification allowed for operation either the 2.4Ghz or 5Ghz bands, however it was/is rarely used in the latter band as it would exclude any clients which could not operate at 5Ghz and the range was lower at 5Ghz (though less congestion). 

802.11ac aka Wi-Fi 5 (2013)

802.11ac didn't introduce many new technologies but provided improved speed and performance by increasing the limits of various features which were in 802.11n, notably:

  • Support for 80Mhz wide channels was mandated (double the 40Mhz of 802.11n).  Wider channels provide greater capacity though, in highly congested areas, it can be counterproductive - think of it as an extra-wide load straddling multiple lanes on a highway.
  • Support for up to 256-QAM  (vs. 64-QAM in 802.11n). The number of QAM carriers multiples the bitrate so you can carry more data within a given channel allocation. You can learn more about QAM in my previous article.

With vendors anxious to release products the first ('Draft 3.0') version of 802.11ac was released in 2013. A later release in 2016 (called Wave 2 by the Wi-Fi Alliance) included 160Mhz channel width support, up to 4 spatial streams (vs.3) and MU-MIMO - all being optional features.

160Mhz channels might be best suited to rural environments or where there are no neighbours, it takes up so much of the band you may see better performance dropping back to narrower channel widths.  That said, in 2020, Ofcom (UK) enlarged the band space to include channels 147-167 allow more channels for general use. You can see a full list of available 5Ghz channels here.

One new feature in 802.11ac was MU-MIMO which I covered in this previous blog.

An important difference between 802.11n (Wi-Fi 4) and 802.11ac (Wi-Fi 5) was that 802.11ac uses the 5Ghz frequency band.  The higher frequency allows for great bit rates and that band is also far less congested, seeing as almost every other Wi-Fi base and client were using the 2.4Ghz band at that point.  Higher frequencies do propagate less far though - they have lower range so 802.11ac may have a lower range than 802.11n but is still sufficient for most typical applications.   You can learn more about the 2.4Ghz vs. 5Ghz range here.

802.11ad & 802.11ay aka 'WiGig'

802.11ad and its successor 802.11ay are unlike the other 802.11 standards in that they are intended for a very short distance, very high-speed applications and use the 60Ghz band. They are not Wi-Fi protocols, and you are unlikely to see them on any consumer devices (yet). I will cover these in more depth in a future article.

802.11ax aka 'Wi-Fi 6'

802.11ax is the latest Wi-Fi standard. It is's the first general use version able to support true Gigabit speed (in ideal circumstances). It introduces various new technologies and enhances the previous ones. I explain the technical improvements 802.11ax (Wi-Fi 6) in my detailed article here.

As always, I hope you find these blogs useful - please do share them using the links above, make comments below and let us know if you have any suggestions for new blog entries.


WiFi 6
Wi-Fi 5
Wi-Fi 4
Evolution of WiFi
Wireless Lan
WiFi Alliance